feat: implement Minimax provider using Anthropic SDK

- Replaced OpenAI SDK with Anthropic SDK for Minimax provider - Updated MiniMaxHandler to extend BaseProvider and implement Anthropic message streaming - Added support for prompt caching when available - Updated tests to mock Anthropic SDK instead of OpenAI - Maintained backward compatibility with existing API Addresses #9082
2026-09-07 08:26:51 +00:00 · 2025-11-06 20:55:56 +00:00 · 2025-11-06 20:55:56 +00:00 · a8829f4f53
commit a8829f4f53
parent bf048492fb
2 changed files with 419 additions and 36 deletions
--- a/src/api/providers/tests/minimax.spec.ts
+++ b/src/api/providers/tests/minimax.spec.ts
@ -8,27 +8,35 @@ vitest.mock("vscode", () => ({
 	},
 }))

-import OpenAI from "openai"
 import { Anthropic } from "@anthropic-ai/sdk"

 import { type MinimaxModelId, minimaxDefaultModelId, minimaxModels } from "@roo-code/types"

 import { MiniMaxHandler } from "../minimax"

-vitest.mock("openai", () => {
+vitest.mock("@anthropic-ai/sdk", () => {
 	const createMock = vitest.fn()
+	const countTokensMock = vitest.fn()
 	return {
-		default: vitest.fn(() => ({ chat: { completions: { create: createMock } } })),
+		Anthropic: vitest.fn(() => ({
+			messages: {
+				create: createMock,
+				countTokens: countTokensMock,
+			},
+		})),
 	}
 })

 describe("MiniMaxHandler", () => {
 	let handler: MiniMaxHandler
 	let mockCreate: any
+	let mockCountTokens: any

 	beforeEach(() => {
 		vitest.clearAllMocks()
-		mockCreate = (OpenAI as unknown as any)().chat.completions.create
+		const mockClient = (Anthropic as unknown as any)()
+		mockCreate = mockClient.messages.create
+		mockCountTokens = mockClient.messages.countTokens
 	})

 	describe("International MiniMax (default)", () => {
@ -41,7 +49,7 @@ describe("MiniMaxHandler", () => {

 		it("should use the correct international MiniMax base URL by default", () => {
 			new MiniMaxHandler({ minimaxApiKey: "test-minimax-api-key" })
-			expect(OpenAI).toHaveBeenCalledWith(
+			expect(Anthropic).toHaveBeenCalledWith(
 				expect.objectContaining({
 					baseURL: "https://api.minimax.io/v1",
 				}),
@ -51,7 +59,7 @@ describe("MiniMaxHandler", () => {
 		it("should use the provided API key", () => {
 			const minimaxApiKey = "test-minimax-api-key"
 			new MiniMaxHandler({ minimaxApiKey })
-			expect(OpenAI).toHaveBeenCalledWith(expect.objectContaining({ apiKey: minimaxApiKey }))
+			expect(Anthropic).toHaveBeenCalledWith(expect.objectContaining({ apiKey: minimaxApiKey }))
 		})

 		it("should return default model when no model is specified", () => {
@ -117,13 +125,13 @@ describe("MiniMaxHandler", () => {
 				minimaxApiKey: "test-minimax-api-key",
 				minimaxBaseUrl: "https://api.minimaxi.com/v1",
 			})
-			expect(OpenAI).toHaveBeenCalledWith(expect.objectContaining({ baseURL: "https://api.minimaxi.com/v1" }))
+			expect(Anthropic).toHaveBeenCalledWith(expect.objectContaining({ baseURL: "https://api.minimaxi.com/v1" }))
 		})

 		it("should use the provided API key for China", () => {
 			const minimaxApiKey = "test-minimax-api-key"
 			new MiniMaxHandler({ minimaxApiKey, minimaxBaseUrl: "https://api.minimaxi.com/v1" })
-			expect(OpenAI).toHaveBeenCalledWith(expect.objectContaining({ apiKey: minimaxApiKey }))
+			expect(Anthropic).toHaveBeenCalledWith(expect.objectContaining({ apiKey: minimaxApiKey }))
 		})

 		it("should return default model when no model is specified", () => {
@ -136,7 +144,7 @@ describe("MiniMaxHandler", () => {
 	describe("Default behavior", () => {
 		it("should default to international base URL when none is specified", () => {
 			const handlerDefault = new MiniMaxHandler({ minimaxApiKey: "test-minimax-api-key" })
-			expect(OpenAI).toHaveBeenCalledWith(
+			expect(Anthropic).toHaveBeenCalledWith(
 				expect.objectContaining({
 					baseURL: "https://api.minimax.io/v1",
 				}),
@ -152,6 +160,10 @@ describe("MiniMaxHandler", () => {
 			const model = handlerDefault.getModel()
 			expect(model.id).toBe("MiniMax-M2")
 		})
+
+		it("should throw error when API key is not provided", () => {
+			expect(() => new MiniMaxHandler({} as any)).toThrow("MiniMax API key is required")
+		})
 	})

 	describe("API Methods", () => {
@ -161,15 +173,24 @@ describe("MiniMaxHandler", () => {

 		it("completePrompt method should return text from MiniMax API", async () => {
 			const expectedResponse = "This is a test response from MiniMax"
-			mockCreate.mockResolvedValueOnce({ choices: [{ message: { content: expectedResponse } }] })
+			mockCreate.mockResolvedValueOnce({
+				content: [{ type: "text", text: expectedResponse }],
+			})
 			const result = await handler.completePrompt("test prompt")
 			expect(result).toBe(expectedResponse)
+			expect(mockCreate).toHaveBeenCalledWith({
+				model: "MiniMax-M2",
+				max_tokens: 16384,
+				temperature: 1.0,
+				messages: [{ role: "user", content: "test prompt" }],
+				stream: false,
+			})
 		})

 		it("should handle errors in completePrompt", async () => {
 			const errorMessage = "MiniMax API error"
 			mockCreate.mockRejectedValueOnce(new Error(errorMessage))
-			await expect(handler.completePrompt("test prompt")).rejects.toThrow()
+			await expect(handler.completePrompt("test prompt")).rejects.toThrow(errorMessage)
 		})

 		it("createMessage should yield text content from stream", async () => {
@ -182,7 +203,38 @@ describe("MiniMaxHandler", () => {
 							.fn()
 							.mockResolvedValueOnce({
 								done: false,
-								value: { choices: [{ delta: { content: testContent } }] },
+								value: {
+									type: "content_block_start",
+									index: 0,
+									content_block: { type: "text", text: testContent },
+								},
+							})
+							.mockResolvedValueOnce({ done: true }),
+					}),
+				}
+			})
+
+			const stream = handler.createMessage("system prompt", [])
+			const firstChunk = await stream.next()
+
+			expect(firstChunk.done).toBe(false)
+			expect(firstChunk.value).toEqual({ type: "text", text: testContent })
+		})
+
+		it("createMessage should handle text delta chunks", async () => {
+			const testContent = "streaming text"
+
+			mockCreate.mockImplementationOnce(() => {
+				return {
+					[Symbol.asyncIterator]: () => ({
+						next: vitest
+							.fn()
+							.mockResolvedValueOnce({
+								done: false,
+								value: {
+									type: "content_block_delta",
+									delta: { type: "text_delta", text: testContent },
+								},
 							})
 							.mockResolvedValueOnce({ done: true }),
 					}),
@ -205,8 +257,15 @@ describe("MiniMaxHandler", () => {
 							.mockResolvedValueOnce({
 								done: false,
 								value: {
-									choices: [{ delta: {} }],
-									usage: { prompt_tokens: 10, completion_tokens: 20 },
+									type: "message_start",
+									message: {
+										usage: {
+											input_tokens: 10,
+											output_tokens: 20,
+											cache_creation_input_tokens: 5,
+											cache_read_input_tokens: 3,
+										},
+									},
 								},
 							})
 							.mockResolvedValueOnce({ done: true }),
@ -218,12 +277,17 @@ describe("MiniMaxHandler", () => {
 			const firstChunk = await stream.next()

 			expect(firstChunk.done).toBe(false)
-			expect(firstChunk.value).toEqual({ type: "usage", inputTokens: 10, outputTokens: 20 })
+			expect(firstChunk.value).toEqual({
+				type: "usage",
+				inputTokens: 10,
+				outputTokens: 20,
+				cacheWriteTokens: 5,
+				cacheReadTokens: 3,
+			})
 		})

-		it("createMessage should pass correct parameters to MiniMax client", async () => {
+		it("createMessage should pass correct parameters to MiniMax client with prompt caching", async () => {
 			const modelId: MinimaxModelId = "MiniMax-M2"
-			const modelInfo = minimaxModels[modelId]
 			const handlerWithModel = new MiniMaxHandler({
 				apiModelId: modelId,
 				minimaxApiKey: "test-minimax-api-key",
@ -248,17 +312,47 @@ describe("MiniMaxHandler", () => {
 			expect(mockCreate).toHaveBeenCalledWith(
 				expect.objectContaining({
 					model: modelId,
-					max_tokens: Math.min(modelInfo.maxTokens, Math.ceil(modelInfo.contextWindow * 0.2)),
-					temperature: 1,
-					messages: expect.arrayContaining([{ role: "system", content: systemPrompt }]),
+					max_tokens: expect.any(Number),
+					temperature: 1.0,
+					system: [{ text: systemPrompt, type: "text", cache_control: { type: "ephemeral" } }],
+					messages: expect.any(Array),
 					stream: true,
-					stream_options: { include_usage: true },
 				}),
-				undefined,
+				expect.objectContaining({
+					headers: { "anthropic-beta": "prompt-caching-2024-07-31" },
+				}),
 			)
 		})

-		it("should use temperature 1 by default", async () => {
+		it("createMessage should handle reasoning/thinking blocks", async () => {
+			const testThinking = "Let me think about this..."
+
+			mockCreate.mockImplementationOnce(() => {
+				return {
+					[Symbol.asyncIterator]: () => ({
+						next: vitest
+							.fn()
+							.mockResolvedValueOnce({
+								done: false,
+								value: {
+									type: "content_block_start",
+									index: 0,
+									content_block: { type: "thinking", thinking: testThinking },
+								},
+							})
+							.mockResolvedValueOnce({ done: true }),
+					}),
+				}
+			})
+
+			const stream = handler.createMessage("system prompt", [])
+			const firstChunk = await stream.next()
+
+			expect(firstChunk.done).toBe(false)
+			expect(firstChunk.value).toEqual({ type: "reasoning", text: testThinking })
+		})
+
+		it("should use temperature 1.0 by default", async () => {
 			mockCreate.mockImplementationOnce(() => {
 				return {
 					[Symbol.asyncIterator]: () => ({
@ -274,11 +368,59 @@ describe("MiniMaxHandler", () => {

 			expect(mockCreate).toHaveBeenCalledWith(
 				expect.objectContaining({
-					temperature: 1,
+					temperature: 1.0,
 				}),
-				undefined,
+				expect.any(Object),
 			)
 		})
+
+		it("should use custom temperature when provided", async () => {
+			const customTemperature = 0.7
+			const handlerWithTemp = new MiniMaxHandler({
+				minimaxApiKey: "test-minimax-api-key",
+				modelTemperature: customTemperature,
+			})
+
+			mockCreate.mockImplementationOnce(() => {
+				return {
+					[Symbol.asyncIterator]: () => ({
+						async next() {
+							return { done: true }
+						},
+					}),
+				}
+			})
+
+			const messageGenerator = handlerWithTemp.createMessage("test", [])
+			await messageGenerator.next()
+
+			expect(mockCreate).toHaveBeenCalledWith(
+				expect.objectContaining({
+					temperature: customTemperature,
+				}),
+				expect.any(Object),
+			)
+		})
+
+		it("countTokens should try API first then fallback to tiktoken", async () => {
+			const content = [{ type: "text", text: "test content" }] as Anthropic.Messages.ContentBlockParam[]
+
+			// First test successful API response
+			mockCountTokens.mockResolvedValueOnce({ input_tokens: 42 })
+			let result = await handler.countTokens(content)
+			expect(result).toBe(42)
+			expect(mockCountTokens).toHaveBeenCalledWith({
+				model: "MiniMax-M2",
+				messages: [{ role: "user", content }],
+			})
+
+			// Then test API failure with fallback
+			mockCountTokens.mockRejectedValueOnce(new Error("Not supported"))
+			result = await handler.countTokens(content)
+			// Should return a number (tiktoken estimate), exact value depends on tokenizer
+			expect(typeof result).toBe("number")
+			expect(result).toBeGreaterThan(0)
+		})
 	})

 	describe("Model Configuration", () => {
--- a/src/api/providers/minimax.ts
+++ b/src/api/providers/minimax.ts
@ -1,19 +1,260 @@
-import { type MinimaxModelId, minimaxDefaultModelId, minimaxModels } from "@roo-code/types"
+import { Anthropic } from "@anthropic-ai/sdk"
+import { Stream as AnthropicStream } from "@anthropic-ai/sdk/streaming"
+import { CacheControlEphemeral } from "@anthropic-ai/sdk/resources"
+
+import {
+	type ModelInfo,
+	type MinimaxModelId,
+	minimaxDefaultModelId,
+	minimaxModels,
+	MINIMAX_DEFAULT_TEMPERATURE,
+} from "@roo-code/types"

 import type { ApiHandlerOptions } from "../../shared/api"

-import { BaseOpenAiCompatibleProvider } from "./base-openai-compatible-provider"
+import { ApiStream } from "../transform/stream"
+import { getModelParams } from "../transform/model-params"
+
+import { BaseProvider } from "./base-provider"
+import type { SingleCompletionHandler, ApiHandlerCreateMessageMetadata } from "../index"
+
+export class MiniMaxHandler extends BaseProvider implements SingleCompletionHandler {
+	private options: ApiHandlerOptions
+	private client: Anthropic

-export class MiniMaxHandler extends BaseOpenAiCompatibleProvider<MinimaxModelId> {
 	constructor(options: ApiHandlerOptions) {
-		super({
-			...options,
-			providerName: "MiniMax",
-			baseURL: options.minimaxBaseUrl ?? "https://api.minimax.io/v1",
-			apiKey: options.minimaxApiKey,
-			defaultProviderModelId: minimaxDefaultModelId,
-			providerModels: minimaxModels,
-			defaultTemperature: 1.0,
+		super()
+		this.options = options
+
+		if (!this.options.minimaxApiKey) {
+			throw new Error("MiniMax API key is required")
+		}
+
+		// MiniMax supports Anthropic-compatible API
+		// https://platform.minimax.io/docs/api-reference/text-anthropic-api
+		this.client = new Anthropic({
+			baseURL: this.options.minimaxBaseUrl || "https://api.minimax.io/v1",
+			apiKey: this.options.minimaxApiKey,
 		})
 	}
+
+	async *createMessage(
+		systemPrompt: string,
+		messages: Anthropic.Messages.MessageParam[],
+		metadata?: ApiHandlerCreateMessageMetadata,
+	): ApiStream {
+		const { id: modelId, info, maxTokens, temperature } = this.getModel()
+		const cacheControl: CacheControlEphemeral = { type: "ephemeral" }
+
+		// Check if the model supports prompt caching
+		const supportsPromptCache = info.supportsPromptCache ?? false
+		const betas: string[] = []
+
+		if (supportsPromptCache) {
+			betas.push("prompt-caching-2024-07-31")
+		}
+
+		let stream: AnthropicStream<Anthropic.Messages.RawMessageStreamEvent>
+
+		if (supportsPromptCache) {
+			// With prompt caching support, handle cache control for user messages
+			const userMsgIndices = messages.reduce(
+				(acc, msg, index) => (msg.role === "user" ? [...acc, index] : acc),
+				[] as number[],
+			)
+
+			const lastUserMsgIndex = userMsgIndices[userMsgIndices.length - 1] ?? -1
+			const secondLastMsgUserIndex = userMsgIndices[userMsgIndices.length - 2] ?? -1
+
+			stream = await this.client.messages.create(
+				{
+					model: modelId,
+					max_tokens: maxTokens ?? 16384,
+					temperature,
+					// Setting cache breakpoint for system prompt so new tasks can reuse it
+					system: [{ text: systemPrompt, type: "text", cache_control: cacheControl }],
+					messages: messages.map((message, index) => {
+						if (index === lastUserMsgIndex || index === secondLastMsgUserIndex) {
+							return {
+								...message,
+								content:
+									typeof message.content === "string"
+										? [{ type: "text", text: message.content, cache_control: cacheControl }]
+										: message.content.map((content, contentIndex) =>
+												contentIndex === message.content.length - 1
+													? { ...content, cache_control: cacheControl }
+													: content,
+											),
+							}
+						}
+						return message
+					}),
+					stream: true,
+				},
+				betas.length > 0 ? { headers: { "anthropic-beta": betas.join(",") } } : undefined,
+			)
+		} else {
+			// Without prompt caching
+			stream = await this.client.messages.create({
+				model: modelId,
+				max_tokens: maxTokens ?? 16384,
+				temperature,
+				system: [{ text: systemPrompt, type: "text" }],
+				messages,
+				stream: true,
+			})
+		}
+
+		let inputTokens = 0
+		let outputTokens = 0
+		let cacheWriteTokens = 0
+		let cacheReadTokens = 0
+
+		for await (const chunk of stream) {
+			switch (chunk.type) {
+				case "message_start": {
+					// Tells us cache reads/writes/input/output
+					const {
+						input_tokens = 0,
+						output_tokens = 0,
+						cache_creation_input_tokens,
+						cache_read_input_tokens,
+					} = chunk.message.usage
+
+					yield {
+						type: "usage",
+						inputTokens: input_tokens,
+						outputTokens: output_tokens,
+						cacheWriteTokens: cache_creation_input_tokens || undefined,
+						cacheReadTokens: cache_read_input_tokens || undefined,
+					}
+
+					inputTokens += input_tokens
+					outputTokens += output_tokens
+					cacheWriteTokens += cache_creation_input_tokens || 0
+					cacheReadTokens += cache_read_input_tokens || 0
+
+					break
+				}
+				case "message_delta":
+					// Tells us output tokens along the way and at the end of the message
+					yield {
+						type: "usage",
+						inputTokens: 0,
+						outputTokens: chunk.usage.output_tokens || 0,
+					}
+					break
+				case "message_stop":
+					// No usage data, just an indicator that the message is done
+					break
+				case "content_block_start":
+					switch (chunk.content_block.type) {
+						case "thinking":
+							// Handle reasoning/thinking blocks if supported
+							if (chunk.index > 0) {
+								yield { type: "reasoning", text: "\n" }
+							}
+							yield { type: "reasoning", text: chunk.content_block.thinking }
+							break
+						case "text":
+							// We may receive multiple text blocks
+							if (chunk.index > 0) {
+								yield { type: "text", text: "\n" }
+							}
+							yield { type: "text", text: chunk.content_block.text }
+							break
+					}
+					break
+				case "content_block_delta":
+					switch (chunk.delta.type) {
+						case "thinking_delta":
+							yield { type: "reasoning", text: chunk.delta.thinking }
+							break
+						case "text_delta":
+							yield { type: "text", text: chunk.delta.text }
+							break
+					}
+					break
+				case "content_block_stop":
+					break
+			}
+		}
+
+		// Calculate total cost if we have usage data
+		if (inputTokens > 0 || outputTokens > 0 || cacheWriteTokens > 0 || cacheReadTokens > 0) {
+			// MiniMax pricing (per million tokens):
+			// Input: $0.3, Output: $1.2, Cache writes: $0.375, Cache reads: $0.03
+			const inputCost = (inputTokens / 1_000_000) * (info.inputPrice || 0)
+			const outputCost = (outputTokens / 1_000_000) * (info.outputPrice || 0)
+			const cacheWriteCost = (cacheWriteTokens / 1_000_000) * (info.cacheWritesPrice || 0)
+			const cacheReadCost = (cacheReadTokens / 1_000_000) * (info.cacheReadsPrice || 0)
+			const totalCost = inputCost + outputCost + cacheWriteCost + cacheReadCost
+
+			yield {
+				type: "usage",
+				inputTokens: 0,
+				outputTokens: 0,
+				totalCost,
+			}
+		}
+	}
+
+	getModel() {
+		const modelId = this.options.apiModelId
+		const id = modelId && modelId in minimaxModels ? (modelId as MinimaxModelId) : minimaxDefaultModelId
+		const info: ModelInfo = minimaxModels[id]
+
+		const params = getModelParams({
+			format: "anthropic",
+			modelId: id,
+			model: info,
+			settings: this.options,
+		})
+
+		return {
+			id,
+			info,
+			...params,
+			temperature: this.options.modelTemperature ?? MINIMAX_DEFAULT_TEMPERATURE,
+		}
+	}
+
+	async completePrompt(prompt: string) {
+		const { id: model, temperature } = this.getModel()
+
+		const message = await this.client.messages.create({
+			model,
+			max_tokens: 16384,
+			temperature,
+			messages: [{ role: "user", content: prompt }],
+			stream: false,
+		})
+
+		const content = message.content.find(({ type }) => type === "text")
+		return content?.type === "text" ? content.text : ""
+	}
+
+	/**
+	 * Counts tokens for the given content using MiniMax's Anthropic-compatible API
+	 * Falls back to tiktoken estimation if the API doesn't support token counting
+	 */
+	override async countTokens(content: Array<Anthropic.Messages.ContentBlockParam>): Promise<number> {
+		try {
+			// Try to use the API's token counting if available
+			// Note: This might not be supported by MiniMax yet
+			const { id: model } = this.getModel()
+
+			// MiniMax might not have token counting endpoint yet
+			// If they add it, it would follow Anthropic's pattern
+			const response = await this.client.messages.countTokens({
+				model,
+				messages: [{ role: "user", content: content }],
+			})
+
+			return response.input_tokens
+		} catch (error) {
+			// Fallback to tiktoken estimation
+			return super.countTokens(content)
+		}
+	}
 }